Možnosti stanovení složek bioplynů, rozpouštědel a degradačních produktů polymerů plynovou

chromatografií s detekcí ve vakuové UV oblastiIng. Tomáš KorbaAMEDIS, Praha

MicroscopyMicrowave Spectroscopy

CrystallographyNMR

HPLC-DAD

FTIRMRI

THz Scanner

XRFImaging

Radiation DetectorNuclear Med

Microwave Oven

The Commercialization of Light

Overview of Molecular Spectroscopy

*Exciting the electron cloud, probing the electronic state of the molecule

0.0

0.5

1.0

1.5

125 175 225Norm

alize

d Abs

orba

nce

Wavelength (nm)

Methane

Vacuum Ultraviolet Detector for GC

100 200 300 400 500 600 700

VUV UV Vis

σà σ*n à σ*

πà π*(isolated π)

πà π*(diene π)

πà π*(conjugated π)

n à π*

Wavelength (nm)

alkanes

unsaturated alkanes

O, N, S, halogens

dienes

aromatics

carbonyls

Schug et al., Anal. Chem. 2014, 86, 8329-8335.

The excitation energies associated with electrons forming most single bonds are sufficiently high that absorption by them is

restricted to the so-called vacuum ultraviolet region (λ<185nm), where components in the atmosphere also absorb strongly. The experimental difficulties associated with the vacuum ultraviolet

are significant; as a result… no further discussion will be devoted to this type of absorption.

Principals of Instrumental Analysis, by Douglas Skoog, Sixth Edition, 2006

VUV: A Gap in The Analytical Toolbox

“The excitation energies associated with electrons forming most single bonds are sufficiently high that absorption occurs in the so-called vacuum ultraviolet region (λ<185nm), where components in the atmosphere also absorb strongly. Because of experimental difficulties associated with the vacuum ultraviolet region, most spectrophotometric investigations of organic compounds have involved longer wavelengths than 185nm.”

Principals of Instrumental Analysis,by Douglas Skoog, Sixth Edition, 2006

Characteristics of VGA-100 (VGA-101)

Schug et al., Anal. Chem. 2014, 86, 8329-8335.

Make-up gas mitigates/alleviates band-broadening

Nondestructive

Full range 120 – 240 nm (430 nm) data acquisition at up to 100 Hz

VUV/UV source lamp (D2); stable signal and robust

Thermostatted flow cell (up to 320 ᵒC (420 ᵒC))

Carrier and make-up gases (He, H2, N2, Ar) largely VUV transparent. No vacuum required!!

Post-run processing for universal or selective detection; library spectral matching; etc.Analyte-specific absorption based on

Beer’s Law principles; Qualitative and quantitative analysis (wide linear range)

Heated transfer line(up to 320 ᵒC (420 ᵒC))

Key Features of GC-VUV

• Universal and selective detection• Unique, class similar spectra• Optical differentiation of isomers• Deconvolution of coeluting

analytes• Automated classification and

speciation of mixtures• Good quantitative performance

Bai et al., J. Chromatogr A 2015, 1388, 244-250.

Permanent gas analysis

Thermal runaway of Li-ion and Li-M batteries

Key Features of GC-VUV

• Aromatics and substituted aromatics

Alkanes vs. Aromatics

Weathered diesel analysis

Bai et al., RevisedNote: The absorbance of e.g benzene at 180 nm is 10 000 more intense than at the typical 254 nm in HPLC

• Universal and selective detection• Unique, class similar spectra• Optical differentiation of isomers• Deconvolution of coeluting analytes• Automated classification and

speciation of mixtures• Good quantitative performance

Library search and Unambiguous Compound Identification

Spectra are very robust;No ghost components in the library match list

Similar But Very Distinct Visual similarities are easily distinguished in the fitting routine; minor differences are significant

The chi-squared distribution is the distribution of a sum of the squares of k independent standard normal random variablesThe chi-squared distribution is used in the common tests for goodness of fit of two criteria of qualitative data,

Optical Differentiation of IsomersUV/Vis spectroscopyAtom connectivity; overlap of molecular orbitalsElectron transition between orbitalsEnergy gap and probability of transition dependent on molecular structure

Diaminobenzene isomers (Handbook of UV Spectra, Chromalytica)

Dimethylnaphthalene isomersHexachlorocyclohexane isomers

Key Features of GC-VUV

• Universal and selective detection• Unique, class similar spectra• Optical differentiation of isomers• (Spectral) Deconvolution of

coeluting analytes• Automated classification and

speciation of mixtures• Good quantitative performance

Polychlorinated biphenyls in Aroclor mixtures

Qiu et al., J. Chromatogr A 2017, 1490, 191-200.

Arochlor 1242

Deconvolution

Beer’s Law: A = ϵbCAtotal = A1 + A2 + … An

Aco-elution = f1A1 + f2A2

Ai = the absorbance of component i at a given

fi = the fit coefficient of component i

Addi

tive

Deco

nvol

ve

• Total absorption is proportional to the product of the concentration and the absorption cross section

• Co-elution is a sum of these products• Linear regression allows for easy deconvolution of the compound

concentrations; even for co-eluting compounds

Spectral Deconvolution of m&p Xylene

(mainlib) o-Xylene

10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 1200

50

100

15 27 29 38

39

41 43 45 49

50

51

5254

6263

64

65

66 74 76

77

78

80 85 8789

91

92

93 97 99 102

103

104

105

106

107108

(mainlib) p-Xylene

10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 1200

50

100

15 1827

29 32 38

39

41 4350

51

5254 57

63 65

66 69 74 76

77

79

80 83 85 87 89

91

92

93 97 102

103

105

106

107

108

(mainlib) Benzene, 1,3-dimethyl-

10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 1200

50

100

15 27 29 38

39

40 43 45 4950

51

52

5462

63

64

65

66 74 76

77

78

80 85 8789

91

92

93 97 99 102

103

104

105

106

107108

o-xylene

p-xylene

m-xylene

Fast GC-VUV Analysis of Terpene IsomersAdditional selectivity which makes chromatographic separation less stringent

Summary:• Terpenes have VUV spectra that are distinct

Ø Includes structural isomers and co-eluting analytes • VUV spectral identification of terpene isomers allows the chromatographic compression of GC

runtimesØ Can shorten GC runtimes by 2 – 3X or more

• Natural and forced co-elutions are deconvolved by VUV softwareØ Eliminates inaccuracy inherent to dropping vertical integration lines to quantitate

Fast GC-VUV Chromatogram of Terpenes VUV Spectral Identification of Monoterpene Isomers

VUV Deconvolution of Co-Eluting Isomers

Key Features of GC-VUV

• Universal and selective detection• Unique, class similar spectra• Optical differentiation of isomers• Deconvolution of coeluting

analytes• Automated classification and

speciation of mixtures (compositional analysis

• Good quantitative performance

Time Interval Deconvolution (TID)

PIONA analysis of finished gasoline (ASTM D8071)

Walsh et al., Anal. Chem. 2016, 88, 11130 – 11138.

PIONA Compound Characterization using VUV PIONA+• Compositional Analysis of gasoline type samples

• Each PIONA compound class displays distinct spectral featuresØ Enables straightforward compound class identification and

quantitation

• PIONA compounds can be speciated through C6, and class identification >C6 using VUV PIONA+

Ø Flexibility to vary chromatographic conditionsØ Specific analytes such as individual oxygenates or aromatics

belonging to the BTEX complex

• ASTM Method D8071 provides information that historically required the use of several ASTM methods or more comprehensive methods.

Ø A single-column GC-VUV separation method with a total runtime of 34 minutes.Ø ASTM approved for PIONA compound analysis in finished gasolineØ No special setups, traps, pre-column tuning, or calibration requirementsØ Automated VUV PIONA+ analysis and reporting

Group plus Compound Identification

Paraffins

Naphthenes Isoparaffins

Aromatics

Olefins Diolefins

Oxygenates

Typical Gasoline Sample

Mass % Report for PIONA and Select Individual Components

Gasoline 87 Octane Number

Key Features of GC-VUV

• Universal and selective detection• Unique, class similar spectra• Complementarity to MS• Deconvolution of coeluting

analytes• Automated classification and

speciation of mixtures• Good quantitative performance

Instrumental Detection Limits

Schug et al., Anal. Chem. 2014, 86, 8329-8335.

Paraffin IDLs averaged 41 pg on column

PAH IDLs averaged 28 pg on column

Terpene IDLs averaged 28 pg on column

FAME IDLs averaged 34 pg on column

Fragrance allergen [A], [B], and [C] IDLs averaged 35, 44, and 36 pg on column

Class 2&3 Residual Solvents200 mg Infant

Acetaminophen2 mL Water

0.25X USP Limit (ppm) Spike

0 2 4 6 8

0.0

0.5

1.0

1.5

2.0

VUV

Abs

orba

nce

Retention Time (min)

140-160 nmEt

hano

l

T-bu

tano

lM

TBE

Hexa

ne2-

Buta

none

& E

thyl

Acet

ate

Cyclo

hexa

neIs

opro

pyl A

ceta

ten-

Hept

ane

Met

hylcy

clohe

xane

Tolu

ene

M &

p-X

ylene

Tetra

lin

Combined Class 2 & 3 Solvents

Analyzing Untargeted Unknowns by GC-VUV: Hangover Pain Relief

Sample spectrum (from Hangover)

Ethanol

Library spectrum (ethanol)

Best Fit w/ VUV Library

Summary: • Applied GC-VUV residual solvents method to over-the-counter hangover medicine• Detected an unexpected compound in the sample injected through static headspace• Identified the unknown analyte as ethanol using VUV spectral library matching

Water Quantitation with GC-VUV

-0.0020.0030.0080.0130.0180.0230.0280.033

125 145 165 185 205 225 245

Nor

mal

ized

Abs

orba

nce

Wavelength (nm)

VUV Absorbance Spectrum Library Match

Measured Absorbance Water Library Spectral Match

Advantages Compared to Karl Fischer:§ Fast and reliable§ No expensive/toxic reagents§ Definitive Quantitation§ Not affected by interferences and side reactions that can

occur using existing KFT methods

Introducing the SVGA-100

Streaming Gas Analysis Instrument

Ethylene / Acetylene Analysis

Ethylene Acetylene

Ethylene in Acetylene – Yes!

Formaldehyde

0.0E+00

1.0E-17

2.0E-17

3.0E-17

4.0E-17

5.0E-17

6.0E-17

115 135 155 175 195 215

Wavelength (nm)C

ross

sec

tion

(cm

-2)

Calculated

Methanol

0.0E+00

4.0E-18

8.0E-18

1.2E-17

1.6E-17

2.0E-17

115 135 155 175 195 215

Wavelength (nm)

Cro

ss s

ectio

n (c

m-2

)

Calculated

Water

0.0E+00

2.0E-18

4.0E-18

6.0E-18

8.0E-18

1.0E-17

1.2E-17

1.4E-17

1.6E-17

115 135 155 175 195 215

Wavelength (nm)

Cro

ss s

ectio

n (c

m-2

) Calculated

Formalin Headspace – No Separation Case• Formalin solution headspace was sampled continuously• Absorption cross sections for water, methanol, and formaldehyde was applied to total

absorption

Additional Thoughts

• No vacuum • Robustness• Reproducibility of spectra• Fast GC

• Spectral vs. chrom resolution

0.00

0.20

0.40

0.60

0.80

1.00

125 145 165 185 205 225

Nor

mal

ized

Abso

rban

ce

Wavelength (nm)

Stearic acid methyl ester

30/12/2016 (SSR = 0.002491)29/01/2019 (SSR = 0.005181)24/02/2017 (SSR = 0.00402)12/04/2017 (SSR = 0.00445)06/05/2017 (SSR = 0.001138)

0.00

0.20

0.40

0.60

0.80

1.00

125 145 165 185 205 225

Nor

mal

ized

Abso

rban

ce

Wavelength (nm)

cis-9,12,15-C18:3

30/12/2016 (SSR = 0.028237)29/01/2017 (SSR = 0.007085)24/02/2017 (SSR = 0.001159)12/04/2017 (SSR = 0.013596)06/05/2017 (SSR = 0.114003)

16

Key Features of GC-VUV

• Universal and selective detection• Unique, class similar spectra• Optical differentiation of isomers• Deconvolution of coeluting

analytes• Automated classification and

speciation of mixtures• Good quantitative performance

What’s Next?• GC-VUV/MS: parallel detection• VUV is complementary to MS

Thank you for your attention!

Science in a new light

korba@amedis.cz

www.vuvanalytics.com